Feature sets for screenshot detection

Loading...
Thumbnail Image
Authors
Sharpe, Lauren
Subjects
Screenshot Detection
Computer Forensics
Triage
Feature Selection
Machine Learning
Bayes
Hough Transform
Entropy
Advisors
Young, Joel
Kolsch, Mathias
Date of Issue
2013-06
Date
Jun-13
Publisher
Monterey, California: Naval Postgraduate School
Language
Abstract
As digital media capacity continues to increase and the cost continues to decrease, digital forensic examiners need progressively more efficient, effective, and tailored tools in order to perform useful media triage. This thesis documents the development of feature sets for classifying images as either screenshots or non-screenshots. Using linear- and intensity-based image information we developed the first (to our knowledge) screenshot detection algorithm. Four feature sets were developed and combinations of these feature sets were tested, with the best results achieving an F-score of 0.98 in ten-fold cross-validation. Requiring less than 0.18 seconds to analyze and classify an image, this is a critical contribution to the state-of-the-art of media forensics.
Type
Description
Department
Computer Science
Organization
Identifiers
NPS Report Number
Sponsors
Funder
Format
Citation
Distribution Statement
Approved for public release; distribution is unlimited.
Rights
This publication is a work of the U.S. Government as defined in Title 17, United States Code, Section 101. Copyright protection is not available for this work in the United States.
Collections